Currently, there are over 5.7 million Americans living with Alzheimer?s disease (AD). Its prevalence and associated costs are projected to triple in coming decades given the aging of the nation?s population. Because of this looming epidemic, delaying the onset of AD symptoms and curbing the progression of the underlying disease process has become a national public health imperative. Delaying symptom onset by as little as five years could reduce the prevalence of AD by half. Unfortunately, currently-available drug treatments for AD are not curative. The urgency of alternative approaches for halting the crisis posed by AD cannot be overstated. There is increasing recognition for the link between physical activity (PA) and prevention of AD. Meta-analytic data of prospective studies have demonstrated that those who are physically active during midlife are at a 35% decreased risk for cognitive decline, and 45% reduced risk for AD in later life. Whereas the beneficial effects of PA on risk for AD have been well studied, investigations of the influence of PA on biomarkers of AD pathology are just beginning. The emerging evidence from our published work and that of other groups suggests that PA is positively associated with many indices of brain health, including cerebral glucose metabolism. Even so, there remain critical gaps in our knowledge. First, the majority of studies have relied of self-report measures of PA which are prone to reporting biases. Second, extant studies have been cross-sectional in nature, and therefore, we currently do not know how baseline levels of PA influence the trajectory of AD biomarkers prospectively. Equally important, we do not know the extent to which changes in PA over time track with changes in AD biomarkers. Therefore, the objective of this proposal is to determine the longitudinal relationship between PA, cerebral glucose metabolism and cognitive function in 50 adults at-risk for AD. We originally studied these individuals in 2013 and now propose to bring them back for a 5-year follow up. Our Central Hypothesis is that longitudinal trajectories of cerebral glucose metabolism and cognition will vary as a function of moderate-intensity PA. We will test our hypothesis with two specific aims: Aim 1: Investigate the longitudinal relationship between PA and cerebral glucose metabolism and Aim 2: Investigate the longitudinal relationship between PA and cognitive function. We predict i) that engagement in moderate-intensity PA performed at baseline (5 years prior) will positively predict longitudinal cerebral glucose metabolism and cognitive trajectories five years later, and ii) changes in moderate-intensity PA will concomitantly associate with changes in cerebral glucose metabolism and cognition over a five year period. This research could have significant translational impact by uncovering a mechanism through which PA curbs the underlining pathophysiology of AD and as a result, delays symptom onset. The project is supported by the extensive resources of the Wisconsin Alzheimer?s Disease Research Center.